Introduction

The demand for antireflective (AR) films has increased with the emergence of a variety of displays in our everyday life because those displays are often mounted on mobile devices that require low energy consumption. Many different types of AR films have been produced using a variety of techniques and materials.1, 2, 3, 4, 5, 6 Despite the different construction and materials, the optical properties of AR films are controlled by only two parameters: the thickness and the refractive index. The strategy of anti-reflection is based on destructive interference between the reflected light at the AR film surface and the interface between the AR and base films. The anti-reflection efficiency is maximized when the film thickness is approximately one-fourth of the wavelength of the light in use. Therefore, AR films are required to have thicknesses of ~100 nm. Among the many different processes used to produce AR films, the wet process, in which AR films are coated by casting solutions onto base films, is preferred for mass production. In the case of a wet process, such as spin coating, control of the thickness of AR films is easily achieved.

In addition, AR films require low refractive indices. In the case of a single-layer AR coating with a refractive index of n between two materials whose refractive indices are n0 and n1, the ideal refractive index n of the AR film is the geometric average of n0 and n1 ().7 If we consider an AR film coated on a common glass (n1=1.5) in air (n0=1.0), the ideal refractive index of the film would be ~1.23. Unfortunately, no existing solid materials have such a low refractive index. The lowest refractive index among commonly known materials is ~1.38 (magnesium fluoride). To reduce the refractive index of a material, one possibility is to make it cellular or porous, where cells and pores are defined as closed and open pores, respectively. In general, cells and pores scatter light and make films opaque; however, if the dimensions of the cells or pores are much smaller than the wavelength of light, the scattering of light from the AR film can be negligible. Therefore, nanocells and nanopores have the potential to reduce the average refractive index of the material toward 1.0 of air, as in Equation (1).

where nfilm is the refractive index of the film, npolym is the refractive index of the copolymer, that is, the refractive index of the non-porous material, nair is the refractive index of air (=1) and p is the porosity of the film. Because nair of unity is lower than the refractive index of any existing material, the larger the porosity in the film is, the lower the refractive index is.

The principal problem is to control the size of the cells or pores that are introduced in the film, of which the thickness should be precisely controlled in the range from 100 to 150 nm. We previously reported the successful introduction of closed cells of a size of tens of nanometers (nanocells) using the block copolymer template assisted with supercritical carbon dioxide (BCTSC) method.8, 9, 10, 11 BCTSC uses supercritical carbon dioxide to introduce cells and pores in polymers similar to microfoams and microcells. Under normal conditions, polymer foaming with CO2 leads to microcells, which are not transparent and are inadequate for optical applications, such as AR coating. Block copolymers self-assemble into a variety of ordered nanostructures, one of which is the spherical structure in the body-centered cubic lattice. The philosophy of BCTSC is to confine the foaming of CO2 in the CO2-philic nanodomains of block copolymers. Some of the successful block copolymers for BCTSC are those with fluorinated polymer blocks, which have high affinity toward CO2. This work was started with a polystyrene (PS-based copolymer in bulk9,11 and thin films,8,10 but we use a poly(methyl methacrylate) (PMMA)-based copolymer for this study because PMMA has a lower refractive index than that of PS (1.60 and 1.49 for PS and PMMA). The block copolymer used in this study is poly(methyl methacrylate-r-glycidylmethacrylate)-b-poly(perfluorooctylethyl methacrylate) (PMGMA-PFMA). The fluorine-containing PFMA block works as a template for CO2-foaming, and the glycidyl group of PMGMA provides cross-linking junctions. We used PMGMA-PFMA to introduce nanocells to reduce the refractive index to 1.25 and to cross-link the nanocellular films so that the cross-linked nanocellular film is chemically robust, even against exposure to the good solvent of PMGMA-PFMA.

Experimental procedure

Synthesis

The synthesis of PMGMA-PFMA was performed by Ru(III)-catalyzed living radical polymerization.12, 13, 14 The polymerization scheme is displayed in Figure 1a. The characteristics of the polymer are listed in Table 1. The non-CO2-philic block is a random copolymer of methyl methacrylate (MMA) and glycidylmethacrylate (GMA). The GMA monomer with epoxy groups was used to introduce cross-linking junctions.

Figure 1
figure 1

Synthesis of poly(methyl methacrylate-r-glycidylmethacrylate)-b-poly(perfluorooctylethyl methacrylate) (PMGMA-PFMA) by Ru(III)-catalyzed living radical polymerization (a) and the schematic images of the nanocellular process (b). The epoxy group of GMA is used to cross-link the film.

Full size image
Table 1 Characteristics of the PMGMA-PFMA block copolymer
Full size table

Film preparation

The films of PMGMA-PFMA were deposited on small pieces of glass or silicon wafer (2–3 cm2), which were used as received. PMGMA-PFMA was dissolved in α,α,α-trifluorotoluene (TFT) in a concentration range of 2–8 wt%, and the solution was spun-cast onto silicon substrates or glass with spin-coating speeds in the range of 2000–4500 r.p.m.

CO2 process

The CO2 process was performed in a high-pressure stainless vessel connected to a high-pressure liquid chromatography pump (PU-2086 plus, JASCO, Tokyo, Japan) mounted with a cooling head to pump condensed CO2. A back pressure controller (FSC-bpg, JASCO) regulated the pressure of CO2 in the vessel. The vessel was placed either in a water bath or an ice bath to control the temperature of the system. The CO2 process consisted of several different steps. During the first step, the film was swollen using high-pressure CO2 (30 MPa) while maintaining the temperature at 45 °C for 45 min. Then, the vessel was placed in an ice bath to lower the temperature to 0 °C while maintaining the pressure at 30 MPa. During the last step of the process, the pressure was lowered at a rate of 0.6 MPa min−1 to 0.1 MPa. The schematic image of the process is shown in Figure 1b.

Cross-linking reaction

The cross-linking reaction was conducted by exposing the nanocellular PMGMA-PFMA thin films to the vapor of a mixture of ethylenediamine and liquid paraffin (50/50 vol%) in a closed glass Petri dish. The cross-linking reaction time was 2 days to ensure complete reaction, and the temperature of the whole system was maintained at 30 °C.

Characterization of film thickness

Film thickness was determined by ellipsometry (M-220, JASCO) using a light wavelength range of 400–800 nm, and an incident angle of 60° with respect to the surface normal.

Characterization of film wettability

The contact angle measurements were conducted using a static mode, with a FACE automatic contact angle meter (CA-V, Kyowa Interface Science, Saitama, Japan) by depositing a drop of the desired liquid (distilled water or hexadecane) on the coated Si substrates. A digital camera automatically detected the shape of droplets and calculated the contact angle of the films.

Surface etching and nanostructure observation

The observation of the embedded nanostructures was performed using a scanning electron microscope (SEM) with a reactive-ion etching using CF4 at a flow rate of 5 ml min−1, a pressure of 10 Pa and a power density of 10 W cm−2. Under these conditions, the etching rate was ~2–3 nm s−1. The exposed internal structures were observed using an SEM (XL20 ESEM-FEG, FEI, Hillsboro, OR, USA) equipped with a field emission gun. A 10-keV electron beam was used for the observation with a working distance of 10 mm. The use of a low-energy electron beam did not cause too much charge up on the surface, so a conductive coating was omitted.

Reflectivity measurement

The glass plates were coated with nanocellular thin films on both sides. The reflectivities of the coated and uncoated glasses were determined with a UV-visible spectrometer (V-550, JASCO) in reflection mode. Incident and specularly reflected lights were −5 and +5 degrees from the normal direction. Reflectivity was calibrated with an aluminum-coated mirror, assuming 90% reflection and dark (0% reflection). The reflectivity of the aluminum mirror was ~90% through the wavelength range in use.

Results and discussion

The spun-cast films on silicon or glass were pressurized using CO2, as described in the Experimental procedure section, to introduce nanocells in a range of diameters from 30 to 70 nm. The successful introduction of nanocells was confirmed by the increased thickness and the reduced refractive index of the film. Because the film is supported on the substrates, the volume change occurs only in the thickness direction. By assuming that the increment of the volume (thickness) is solely due to the volume of the introduced nanocells (voids), the porosity can be estimated as

where ti and tf are the initial and final thicknesses, respectively. By using a pressure of 30 MPa during the CO2 process, the porosity reached 38%, and the refractive index was lowered to 1.26. Then, the films were cross-linked using a vapor of ethylenediamine. A schematic image of the whole nanocellular film fabrication process is shown in Figure 1b. The observation using SEM of the film surface showed sporadic nanopores at the surface of the film (Figure 2a). Between the pores at the surface, the shadow of the nanocells embedded under the surface is visible. To reveal these embedded nanocells, the top 50 nm of the film was etched using reactive-ion etching. After reactive-ion etching, the nanocells were visible all over the film, which justifies the high porosity of the film (Figure 2b). The pore diameter ranges from 30 to 70 nm. The cells are mostly cylindrical and round in shape, and lamellar pores are not observed, although some nanocells are connected to neighboring cells. Because the top 50 nm of the film is less porous than the inner part, reactive-ion etching allowed us to increase the porosity of the film. By doing so, we reduced the refractive index to 1.23, which is more evidence for the presence of dense layer at the surface.

Figure 2
figure 2

The SEM image of a CO2-treated PMGMA-PFMA thin film (a) and the SEM image of the same film after etching (b).

Full size image

To analyze the AR properties of the nanocellular coating, two films were prepared on glass and silicon substrates and were treated similarly. The film on silicon was used to measure the thickness and refractive index, which were assumed to be the same as the film on glass. We fabricated the film with a refractive index of 1.26 and measured the reflectivity of the glass coated with the film using UV-visible spectroscopy in reflection mode. The results of the reflectivity experiments are shown in Figure 3. The excellent anti-reflection property of our single-layer nanocellular coating is clearly demonstrated. At the nominal wavelength, the reflection of glass is lowered from 9.5 to 0.1%. On the basis of the work of Orfanidis,15 we simulated a film with a refractive index of 1.26 and a thickness of 120 nm on a glass substrate. Our experimentally observed reflectivity shows good agreement with the simulated AR film. Therefore, the nanocellular films create films with a low refractive index, as expected from the effective medium approximation.

Figure 3
figure 3

Reflection spectra of bare and nanocellular PMGMA-PFMA-coated glasses compared with the simulated reflectivity of a single-layer film with a refractive index of n=1.26.

Full size image

As suggested previously,8,10 the surfaces of nanocellular films are covered by fluorinated blocks to minimize the interfacial energy. We performed contact angle measurements on such nanocellular films with water and oil (hexadecane). The surface of the nanocellular thin film exhibits high contact angles for water and hexadecane, of 120 and 60 degrees, respectively, and it hence repels water and oil. The surfaces of the nanocellular films have larger contact angles with water than that of polytetrafluoroethylene, as shown in Figure 4a. Such hydrophobic surfaces can be linked to the large amount of fluorocarbons segregated on the surface of the film. The surface of the nanocellular film has a fairly large contact angle against n-hexadecane (oil), with a low surface tension of 27.1 mN m−1. Even oleophobic polytetrafluoroethylene has a n-hexadecane contact angle of 43 degrees.16 The larger contact angle (60 degrees) of the nanocellular film surface can be justified by the existence of the CF3 groups of the side chains of PFMA at the surface that are enhanced by supercritical CO2.17 The fluorine-containing block is shielding the more hydrophilic section of PMGMA from water and hexadecane. Such hydrophobic and oleophobic properties are ideal for anti-reflection coatings to prevent fouling because the anti-reflection coating must be used for the topmost layer.

Figure 4
figure 4

Picture (a) of a water droplet on the surface of a CO2-processed nanocellular PMGMA-PFMA film. Appearance of the original (b, top), nanocellular (b, middle) and cross-linked nanocellular (b, bottom) films after exposure to toluene (b, left) and α,α,α-trifluorotoluene (TFT; b, right). The circular spots are the area exposed to the corresponding solvents. Only the cross-linked nanocellular (bottom) films are intact.

Full size image

The nanocellular film is very stable in air and lasts several years. In addition, the nanocellular thin films are stable in water due to fluorinated blocks. The nanocellular thin films also show unique fracture resistance, but the robustness, especially against exposure to organic solvent, is an anticipated problem. For chemical robustness experiments, we selected toluene and TFT. Those solvents are good solvents for the nanocellular film because toluene is a typical solvent of PMMA and because TFT was the solvent that was used to spin-cast the film. A droplet of each solvent was placed on the surface of the film. After 30 s of exposure, the solvent was blown out with nitrogen gas. Figure 4b shows the traces of the droplets on the film after solvent exposure. We prepared six films with the same initial thickness of 75 nm. Four of the films were treated with CO2 to be nanocellular films with an increased thickness of 120 nm. Two of the four nanocellular films were cross-linked by ethylenediamine vapor. In the case of the as-cast film, the film thickness was reduced from 75 to 55 nm under the droplet of toluene and from 75 to 0 nm under the droplet of TFT. Therefore, toluene and TFT dissolved the films partially and completely, respectively, as indicated by the change of interference color of the films. The thickness of the nanocellular films was originally 110 nm but was decreased by toluene to <65 nm, which is a thickness equivalent or slightly lower than that of an as-cast film (70 nm). TFT dissolved the nanocellular film completely, similar to the as-cast film. Therefore, the nanocellular films lose all of the cells and pores, even when the films survive solvent exposure. In contrast, the resistance of the cross-linked nanocellular films against toluene and even against TFT was excellent. Ellipsometry measurements on the spots revealed no change of thickness. No sign of collapse or dissolution can be observed in Figure 4b. We conclude that the cross-linked nanocellular films have excellent chemical robustness, even against exposure to good solvents.

Conclusion

The nanocellular thin films based on PMGMA-PFMA showed very low refractive indices and, hence, low reflection as AR coatings as well as good chemical robustness and both water- and oil-repellent properties. This novel material using the CO2 process can be exploited as a high-performance coating for AR purposes.